Analysis Main Engine Reverse Failure The seat for one valve, required for the opening of the starting air valve, was found to have deteriorated by some 50percent as compared to original specifications. This degradation of performance reduced the pressure of air delivered to start the engine. When main engine control was transferred to the engine room, the engineers overrode the automated starting sequence and quickly started the engine astern. The duration of the start attempt in this mode is then limited only to the amount of starting air available. An operator can activate the start sequence for a longer period of time than that of the preset limits. In manual mode, the engineers were able to extend the time the compressed air was applied to start the engine. It is probable that, by doing this, they were able to compensate for the valve seal which had deteriorated by some 50percent. Information gathered during the investigation points to a deteriorated valve seat as the most probable cause of the main engine reverse failure when ordered in automatic mode from the bridge at 1337. It has not been established why, approximately 2hours before leaving anchorage, the reverse start as tested from the bridge was successful. Failure to start at 1337 could have been due to deterioration of the valve, its seat or its Oring at that time. Berthing Methods and Speed of Approach The impact speed was at least 3.5to 4knots. At the time of impact, astern power was functioning and both anchors had been deployed. This suggests that the speed of approach just prior to the initial half-astern order was at least 5or 6knots. Additionally, due to the speed of advance of the Utviken, the force exerted by the tug PaulENo.1 (which was not made fast to the vessel) in the last minutes before impact was at an angle of less than perpendicular to the vessel. It follows that the forward component of the resultant pushing force added to the ahead motion of Utviken, against the resistance of the deployed anchors and astern power. Berthing Methods One accepted procedure for berthing at Pier23 involves slowing the vessel's headway almost to a stop at the turning point to line up the vessel for movement into the slip. This allows the tugs to do the work in turning the vessel and proceeding slowly to the berth. The alternative is to make a running entry in one continuous motion from the turning point, through approximately 50degrees, and arrive at the berth with minimal tug assistance. Astern power is put on at a critical point to begin slowing the vessel and initiate the swing to starboard (for right-hand fixed-pitch-propeller-equipped vessels). The latter was apparently the method chosen by the pilot of the Utviken. The time saving between the two berthing methods is estimated to be between 15and 25minutes, which is negligible. While a running entry undoubtably saves some time, the minimum speed necessary for a vessel to maintain steerage way (a necessary condition up to the point where astern power is applied) makes this berthing method a higher risk option. Manoeuvring room forward and to each side of the approach channel is restricted. It is even more restricted, as it was in this case, when another vessel is berthed opposite the entranceway to Pier23. In the event of any mechanical or control failure or a misjudgment by the bridge team, some degree of adverse consequence is almost inevitable. Speed Given the adverse effect on stopping distance of a restricted UKC at this location, it is reasonable to assume a crash stop would produce an advance of approximately 3cables at an initial speed of 4to 6knots. The restricted waters of the southeast harbour and the speed of the Utviken (a result of the berthing method chosen) left little time or room to respond adequately to any unexpected event. Neither anchors nor tugs were used effectively thereafter to prevent the striking under the circumstances. The slower speed, tug-controlled berthing method is used more frequently. In the event of mechanical failures or of difficulty/error in manoeuvring, the bridge team has more time to assess and correct the approach. Additionally, the effectiveness of anchors to stop, reduce speed or assist in turning the vessel is relative to the speed of the vessel over the ground. Effective Use of Tugs Made Fast or not Made Fast It remains at the pilot's discretion whether or not to have the assisting tugs made fast or not. If they are not made fast, however, their usefulness is limited to pushing and they have essentially no ability to slow the vessel. If not made fast, the tug assisting aft can proceed around the stern to either the starboard or port quarter and be effective as long as the vessel is moving ahead at a relatively slow speed. Any increase in speed by the assisted vessel significantly diminishes the effectiveness of tugs not made fast to the vessel. When deploying tugs ... [the ship handler should] always consider not just the basic berthing manoeuvre ... but whether those tugs are in a position to arrest his forward progress in the event of an emergency.4 In the 1993striking of the HamiltonEnergy and the ProvmarTerminal, the fact that the tugs were not secured to the assisted vessel was identified as one of the contributing factors in the accident. Having tugs made fast makes them more effective in slowing or stopping the assisted vessel, or applying a pulling force perpendicular to the vessel, if necessary. On the other hand, an increased speed, such as that necessary for a running entry, puts tugs which are not made fast at risk. If the tugs are not able to check the vessel's speed, the tugs and their crews may be in danger of being crushed between the assisted vessel and another vessel or wharf. Because of this danger, there is a reluctance amongst pilots to make fast the assisting tugs given the restricted manoeuvring room at Pier23. However, the risk to the tugs is minimal, whether made fast or not, when the speed of approach is reduced and the tug-controlled berthing method is used. Given a reduced speed of approach, tugs which are made fast can be used to better advantage. When the running entry method is used, however, the advantages of having tugs make fast appear to be outweighed by the increased risk to the tugs and their crews. Pivot Point, Turning Levers and Moments During the final minutes before the striking, the pilot instructed the lead tug, PaulENo.1, to push with full power on the port bow of the Utviken in order to move the bow to starboard. The PaulENo.1 was the more powerful tug, at 1100hp, and was already positioned at the fore part of the Utviken. The second assisting tug, LacVancouver, was not used at this time. The pivot point of a vessel at rest is approximately amidships. When this same vessel is making headway, the pivot point moves to about one quarter of the ship's length from forward.5 It can thus be appreciated that the turning lever of the forward tug assisting a vessel making headway is reduced and that of the after tug increased. In this occurrence, the unsecured tug was pushing just aft of the break of the forecastle, or approximately 35m from forward. Because the Utviken was making headway, the pivot point of the vessel would have been approximately 50m from the bow, giving a turning lever of some 15m for the forward tug. The aft tug would have had a turning lever in the order of 120m. Even considering the relative difference in tug power (1100hp forward versus 700hpaft), the aft tug pulling on the port side, if made fast or pushing on the starboard side, would have produced about five times the turning moment of the forward tug (seeFigure2). Additionally, once the anchors were let go, the pivot point would have moved even further forward, to a point between the two windlasses. The forward tug's ability to effect a starboard turning moment at this time would have been minimal. Figure2. Pivot points, turning levers and moments In the last minutes before the striking, the aft tug was not used in a manner most effective in turning the vessel to starboard. For the forward tug, the reduced turning lever, the angle of application of the pushing force, and the deployment of the assisted vessel's port anchor all decreased the effectiveness of turning the vessel to starboard. Although the instructions for the forward tug to push full on the port bow were taken in good faith and with the best of intentions, the starboard turning moment applied on the Utviken by this action was minimal. Also, contrary to the wishes of all concerned, the assisted vessel's speed reduction was hampered by the forward component of the pushing forces applied by the PaulENo.1. Deployment of Anchors Immediately after the bridge team realized the main engine was not responding to the second bridge control command for full-astern power, and since the manoeuvring room was rapidly diminishing, the master ordered the starboard anchor let go. In quick succession to the master's order, the pilot, who was also very concerned about the close proximity of the tankers, ordered both port and starboard anchors dropped. The anchor cables were checked and veered alternately. This undoubtedly helped slow the vessel, but since both anchors were deployed, the vessel's ability to turn to starboard was severely reduced. The amount of cable veered also increased the risk of underwater damage to the hull by the flukes of the anchors. A 20 to 25m movement to starboard would have resulted in the Utviken's bow missing the HamiltonEnergy and the ProvmarTerminal and possibly clearing the end of Pier24. The benefit of deploying the starboard anchor on a limited scope of chain, with the port bow tug pushing, would help to create a turning motion to starboard. Bridge Resource Management The master had confidence in the pilot's ability, but also thought the vessel was proceeding too rapidly as the vessel was approaching Pier23. The essential element of good bridge resource management (BRM) is the challenge and response aspect between the authority of the person conning the vessel (in this case, the pilot) and the assertiveness of the other bridge team members. Although the master did not communicate his concerns to the pilot or challenge him as the vessel rapidly bore down on the moored tankers, he took positive action to increase the vessel's rate of turn and to reduce the vessel's speed. Location of the Tankers Pier 24 has been allocated by the Hamilton Port Authority (HPA) to Provmar FuelsInc. It begins at the northwest end of the pier and extends southeasterly for several hundred metres. It is the location of the permanently moored fuel storage vessels ProvmarTerminal and ProvmarTerminalII, as well as the fuelling ship HamiltonEnergy. There are inherent risks associated with this location; this is the second major striking of these vesselssince1993. In a report involving the striking of a permanently moored passenger vessel in St. Louis Harbour, the National Transportation Safety Board in the United States (US) recommended that the owners site the vessel in a place that is protected from waterborne risk events.6 Other authorities, such as the US Coast Guard (CG), cite risk mitigation measures that include location and protective cells for such vessels.7 Also, the USCG has produced a Permanently Moored Vessel Initial Risk Assessment Form. Finally, a report on site selection and risk mitigation for permanently moored vessels was issued in1999 by the USCG. The HPA is aware that the north end of Pier24 poses higher inherent risks than other locations, although a formal risk assessment has never been carried out. After the occurrence, together with the owners of the fuel oil terminal, an informal risk assessment was conducted. Notwithstanding this approach, there are advantages and further benefits by conducting a formal risk assessment versus intuitive, informal or reactive methods. Some of the advantages include the following: traceability of decisions; identification of potential hazards; identification of potential failure modes; quantitative risk statements; identification of important contributors to risk; and risk reduction solutions. Emergency Warning As the Utviken approached the moored vessels, unable to turn sufficiently to starboard or slow to avoid impact, neither the general alarm nor the warning signal was sounded to advise the crew of these vessels of the imminent danger. Most of the crew aboard the Utviken, HamiltonEnergy and ProvmarTerminal were not alerted to any personal danger. Past TSB occurrence investigations have highlighted the importance of warning crews of an impending danger: It appears that there is either an involuntary lapse or a reluctance to sound the general alarm and/or give warning blasts on the whistle during an emergency. The main engine failed to reverse on two attempts from the wheelhouse control. Upon inspection, a faulty pneumatic start system control valve was found and most probably contributed to the failed reverse engine order at 1337. The speed of the Utviken, just prior to the turn to Pier23, left little or no room for error or mechanical failure. Methods employed to avoid the berthed vessels, such as using an unsecured tug pushing forward and the release of both anchors, were ineffective. While it had a limited success in reducing the vessel's speed, the deployment of the port anchor and the length of chain used on both anchors hindered the ability of the vessel to come to starboard.Findings as to Causes and Contributing Factors The main engine failed to reverse on two attempts from the wheelhouse control. Upon inspection, a faulty pneumatic start system control valve was found and most probably contributed to the failed reverse engine order at 1337. The speed of the Utviken, just prior to the turn to Pier23, left little or no room for error or mechanical failure. Methods employed to avoid the berthed vessels, such as using an unsecured tug pushing forward and the release of both anchors, were ineffective. While it had a limited success in reducing the vessel's speed, the deployment of the port anchor and the length of chain used on both anchors hindered the ability of the vessel to come to starboard. An effective BRM environment was not present on the bridge. As such, the master did not feel comfortable challenging the pilot in respect to the vessel's speed just prior to the turn to Pier23. Because of prevailing winds, harbour geography and position relative to Pier23, both the probability and severity of adverse consequences to the permanently berthed vessels at Pier24 are high. The tugs remained unsecured while assisting the Utviken. By remaining unsecured, the tugs could not be used to their full potential in various manoeuvres to assist the berthing vessel. No emergency signal was sounded by the Utviken, nor was a warning given aboard the moored vessels, prior to impact.Findings as to Risk An effective BRM environment was not present on the bridge. As such, the master did not feel comfortable challenging the pilot in respect to the vessel's speed just prior to the turn to Pier23. Because of prevailing winds, harbour geography and position relative to Pier23, both the probability and severity of adverse consequences to the permanently berthed vessels at Pier24 are high. The tugs remained unsecured while assisting the Utviken. By remaining unsecured, the tugs could not be used to their full potential in various manoeuvres to assist the berthing vessel. No emergency signal was sounded by the Utviken, nor was a warning given aboard the moored vessels, prior to impact. Safety Action Action Taken Vessel Approach to Pier 23, Hamilton Harbour The Hamilton Port Authority (HPA) has drafted new practices and procedures concerning, interalia . . . every vessel that is destined for Piers23 to26, once they clear the Burlington Canal and make the turn toward Pier23, is to test the ship's ability to go astern. Location of the Tankers at Pier 24 As a result of this and a previous occurrence, the owners of Provmar Fuels Inc., together with the HPA, have moved the terminal tanker vessels to the southeast, giving approximately 45m of clear dock between the vessels and the northeast corner of Pier24. In addition negotiations are ongoing to remove the two tankers from their present location and to replace their storage capacity with fuel tanks ashore. Tug Assistance The Hamilton Port Authority has directed that vessels proceeding to Pier23 shall use three tugs to assist, until new harbour regulations for vessels proceeding to Pier23 are in place. The Great Lakes Pilotage Authority will take action to ensure proper powered tugs will be available for a better towing assistance in Hamilton Harbour and will also in consultation with the Port Authority and the pilots look at the feasibility and safety of having the tugs tethered versus untethered. Vessel Management Action M/V Utviken and the sister vessel M/V Inviken now carry out all manoeuvring of the main engines from the engine room while on the Great Lakes. Bridge Resource Management Training The Great Lakes Pilotage Authority advises that, while the authority requires all its pilots to attend bridge resource management (BRM) training courses, it recognizes the need for supplementary courses to continuously support BRM principles. In its five-year training program, the Authority is reviewing Bridge Team Management courses and Full Mission Simulation programs that would incorporate and reinforce BRM practices and procedures.